JP2020057206A - Information processing device - Google Patents

Abstract

To present an item corresponding to preference of a user even when information for estimating the preference of the user is insufficient.SOLUTION: An attribute estimation unit 101 estimates attribute of each SNS user from posting history of the plurality of SNS users. A preference estimation unit 102 estimates preference of each SNS user from the posting history of the plurality of SNS users. A storage unit 103 associates and stores the estimated attribute and the preference. A specification unit 104 specifies, for a customer user to whom an item to be recommended, the preference corresponding to the attribute of the user from storage contents of the storage unit 103.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an information processing device.

  There is a mechanism for recommending goods and services (hereinafter, these are collectively referred to as items). For example, Patent Document 1 proposes a mechanism for recommending a product based on the reputation of the product in an SNS (Social Networking Service).

JP 2014-67359 A

  An object of the present invention is to present an item corresponding to a certain user's preference even when information for estimating the preference of a certain user is insufficient.

  The information processing apparatus according to claim 1, wherein attribute estimation means for estimating attributes of each of the users from the posting histories of the plurality of users, and estimating preferences of each of the users from the posting histories of the plurality of users. A preference estimating unit, a storage unit for storing the estimated attribute and the preference in association with each other, a specifying unit for specifying a preference corresponding to an attribute of a user as a recommendation partner from the storage content of the storage unit, A presentation unit that presents an item corresponding to the preference specified for the partner user to the recommended partner user.

  In the information processing apparatus according to claim 2, in the configuration according to claim 1, when the preference estimating unit estimates a preference from a posting history that matches a predetermined condition, the preference estimating unit determines a relationship between the posting history and the preference. The user's preference is estimated by performing weighting.

  The information processing apparatus according to claim 3, in the configuration according to claim 2, wherein the preference estimating means is configured such that when the user who posted is not the original creator of the posting history, The weight for the relationship between the posting history and the preference is made smaller than in the case of the original creator.

  According to a fourth aspect of the present invention, in the configuration according to the second aspect, the preference estimating unit is configured such that the posting history for the purpose of advertisement or promotion is more compared to the posting history for purposes other than advertising or promotion. The weight for the relationship between the posting history and the preference is reduced.

  In the information processing apparatus according to claim 5, in the configuration according to claim 1, the preference estimating unit determines a relationship between the posting history of the user and the preference according to a user attribute estimated by the attribute estimating unit. Change the weight for.

  In the information processing device according to claim 6, in the configuration according to claim 5, the preference estimating unit determines the posting history of the user and the preference according to a type of the attribute of the user estimated by the attribute estimating unit. Change the weight for the relationship.

  In the information processing apparatus according to claim 7, in the configuration according to claim 5, the preference estimating means includes a posting history of the user and the preference in accordance with estimation accuracy of a user attribute estimated by the attribute estimating means. Change the weight for the relationship with.

  In the information processing apparatus according to claim 8, in the configuration according to claim 7, the preference estimating unit adds an attribute other than the attribute of the user as the recommendation partner, and compares the posting history of the user with the preference. Change the weight for the relationship.

  In the information processing apparatus according to claim 9, in the configuration according to claim 7, the preference estimating unit takes into account the posting history of the user in consideration of a probability that the attribute is an attribute other than the attribute estimated by the attribute estimating unit. The weight for the relationship with the preference is changed.

  In the information processing apparatus according to claim 10, in the configuration according to claim 1, if the attribute that is the same as the attribute of the user as the recommendation partner is not stored in the storage unit, another attribute is set. The preference of the user to be the recommended partner is specified using the preference of the user.

In the information processing device according to claim 11, in the configuration according to claim 10, when the identification unit does not store the same first attribute as the attribute of the user as the recommendation partner in the storage unit, The information processing apparatus according to claim 10, wherein a preference corresponding to a second attribute of the attribute of the user as the recommendation partner is specified.

  In the information processing apparatus according to claim 12, in the configuration according to claim 10, when the identification unit does not store the same attribute as the first attribute of the user as the recommendation partner in the storage unit, According to the user's preference of the second attribute having a preference similar to the preference of the user as the recommendation partner, the preference corresponding to the same second attribute as the attribute of the user as the recommendation partner is specified.

  In the information processing apparatus according to a thirteenth aspect, in the configuration according to the twelfth aspect, an attribute of a user having a preference similar to the preference of the user as the recommendation partner is determined in advance.

  According to a fourteenth aspect of the present invention, in the configuration according to the twelfth aspect, the attribute of the user having a preference similar to the preference of the user as the recommendation partner is determined according to the numerical value of the attribute or the proximity of the position. Have been.

  In the information processing apparatus according to claim 15, in the configuration according to claim 12, the preference of the user as the recommendation partner for the first item is determined according to the preference of another user for the second item.

According to the information processing apparatus of the first aspect, even if information for estimating a certain user's preference is insufficient, an item corresponding to the certain user's preference is presented.
According to the information processing apparatus described in claim 2, the preference of the user is estimated in consideration of the relationship between the posting history and the preference.
According to the information processing apparatus of the third aspect, the weight for the relationship between the posting history and the preference can be changed depending on whether the posting user is not the original creator of the posting history.
According to the information processing apparatus of the fourth aspect, the weight of the relationship between the posting history and the preference can be changed between the posting history for the purpose of the advertisement or the advertisement and the posting for the purpose other than the advertisement or the advertisement.
According to the information processing apparatus of the fifth aspect, the weight for the relationship between the posting history and the preference of the user can be changed according to the attribute of the user.
According to the information processing apparatus of the sixth aspect, the weight for the relationship between the posting history and the preference of the user can be changed according to the type of the attribute of the user.
According to the information processing apparatus of the seventh aspect, the weight for the relationship between the posting history and the preference of the user can be changed according to the estimation accuracy of the attribute of the user.
According to the information processing apparatus of the eighth aspect, the weight for the relationship between the posting history and the preference of the user can be changed by taking into account attributes other than the attribute of the user as the recommendation partner.
According to the information processing apparatus of the ninth aspect, the weight for the relationship between the posting history and the preference of the user can be changed in consideration of the probability that the attribute is an attribute other than the attribute estimated by the attribute estimating unit.
According to the information processing apparatus of the tenth aspect, even when the storage unit does not store the same attribute as the attribute of the user to be recommended, the preference of the user to be recommended is specified.
According to the information processing apparatus of the present invention, even when the storage unit does not store the same first attribute as the attribute of the user to be recommended, the preference of the user to be recommended is specified.
According to the information processing apparatus of the twelfth aspect, even when the storage unit does not store the same first attribute as the attribute of the user to be recommended, the preference of the user to be recommended is specified.
According to the information processing apparatus of the thirteenth aspect, the attribute of the user having a preference similar to the preference of the user to be recommended is determined in advance.
According to the information processing apparatus of the present invention, the attribute of the user having a preference similar to the preference of the user to be recommended is determined according to the numerical value of the attribute or the proximity of the position.
According to the information processing apparatus of the present invention, the preference of the user as the recommendation partner with respect to the first item is determined according to the preference of another user with respect to the second item.

1 is a block diagram illustrating a configuration of an information processing system according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating a hardware configuration of the information processing apparatus. FIG. 2 is a block diagram illustrating a functional configuration of the information processing apparatus. 6 is a flowchart illustrating an operation of the information processing device.

<Structure>
Hereinafter, embodiments of the present invention will be described. FIG. 1 is a block diagram illustrating a configuration of an information processing system 10 according to the embodiment of the present invention. The information processing system 10 includes an information processing apparatus 1, a plurality of posting sites 2a and 2b (hereinafter referred to as posting sites 2), and a network 3 such as the Internet for connecting these so that they can communicate with each other. The posting site 2 is, for example, a computer that provides services called SNS (Social Networking Service), blog, and bulletin board. The posting site 2 stores data posted by a user using the SNS (hereinafter, referred to as an SNS user) (hereinafter, this data group is referred to as a posting history), and provides this to a device connected to the network 3. . The information processing device 1 estimates attributes such as the age and gender of the SNS user and the degree of preference of the SNS user for the item from the posting history stored in the posting site 2 and specifies the relationship between the two. I do. Then, the information processing device 1 uses this relationship to specify an item to be recommended for a target user (hereinafter, referred to as a customer user) to be a recommendation partner based on the attribute of the customer user. Present information about the item to the customer user.

  FIG. 2 is a block diagram illustrating a hardware configuration of the information processing apparatus 1. The information processing device 1 includes a control unit 11, a communication unit 13, and a storage unit 12. The control unit 11 includes an arithmetic processing device such as a CPU (Central Processing Unit) and a storage device such as a ROM (Read Only Memory) and a RAM (Random Access Memory). The ROM stores a program to be executed by the CPU. The CPU controls the operation of each unit of the information processing apparatus 1 by executing a program stored in the ROM or the storage unit 12 using the RAM as a work area. The communication unit 13 includes a communication circuit and performs data communication. The storage unit 12 is a storage unit such as a hard disk, for example, and stores a data group and a program group used by the control unit 11.

  FIG. 3 is a block diagram illustrating a functional configuration of the information processing device 1. The acquisition unit 100 acquires a posting history from each posting site 2 via the network 3.

  The attribute estimating unit 101 estimates attributes of each SNS user from the posting histories of a plurality of SNS users. Specifically, for example, the attribute estimating unit 101 machine-learns the relationship between the text data included in the post and the attribute of the SNS user in advance using a technique such as supervised learning. For example, when a word frequently used by women in their twenties frequently appears in a post, the attribute estimating unit 101 learns the relationship that the SNS user who posted the word is a woman in his twenties. Using the learning model, the attribute estimating unit 101 estimates, for each post, the attribute of the SNS user who made the post. The attributes of the SNS user here include, for example, gender, age (or age), occupation, place of residence, and the like.

  The preference estimating unit 102 estimates the preference of each SNS user from the posting histories of the plurality of SNS users. More specifically, the preference estimating unit 102 uses a method (e.g., sentiment analysis) that analyzes the SNS user's emotions regarding the item using natural language processing, for example. Of the preference of the SNS user who performed the above is estimated. In other words, the posts that are subject to the preference estimation by the preference estimation unit 102 are limited to posts that include a word meaning an item.

  The storage unit 103 stores the estimated attributes and preferences in association with each other. For example, the storage unit 103 statistically processes the relationship between the attribute (for example, a woman in her twenties) and the preference (for example, a product type, a place name as a travel destination candidate, and a cigarette brand) estimated for each post for each post. And then store the result. For example, as a result of the processing of each SNS user, “attribute: 20s”, “sex: female”, “item: Hawaii”, “preference: positive” and the like are stored.

  The specifying unit 104 specifies a preference corresponding to an attribute of the customer user from a storage content of the storage unit 103 for a certain customer user who is a recommendation partner of the item. Specifically, the specifying unit 104 specifies an attribute of a customer user to be an item recommendation partner in a customer list or the like including at least the user identification information and the attribute of the customer user, and further specifies the attribute of the customer user. The corresponding preference is specified from the storage content of the storage unit 103. For example, if the attribute of the customer user to be recommended is “female in their 20s”, the content such as “item: Hawaii” and “favorite: positive” is specified as the item and the corresponding preference.

  The presentation unit 105 presents an item corresponding to the preference specified for the customer user as the recommendation partner to the customer user. Specifically, the presenting unit 105 includes a specifying unit in an item list or the like including at least information on items that match the taste of the SNS user and items that match the tastes (such as item names and explanation information that describes the items). Information about the item corresponding to the preference specified by 104 is transmitted to the communication device of the user specified by the user identification information and displayed. For example, if the SNS user's preference of a woman in her twenties is “positive” for “Hawaii” and a list of travel destinations is given as an item, and the item name contained therein is “Hawaii”, The item is transmitted as a recommended item to the communication device or the like of the customer user.

<Operation>
FIG. 4 is a flowchart showing the operation of the information processing apparatus 1. First, the acquisition unit 100 acquires a plurality of posting histories from each posting site 2 via the network 3 (Step S11).

  Next, the attribute estimating unit 101 estimates attributes of each SNS user from the posting histories of the plurality of SNS users (step S12).

  The preference estimating unit 102 estimates the preference of each SNS user from the posting histories of the plurality of SNS users (step S13).

  The storage unit 103 stores the estimated attributes and preferences in association with each other (step S14).

  The specification unit 104 specifies a customer user who is a recommendation partner of the item, which is a recommendation partner of the item, in a customer list or the like (step S15).

  The specifying unit 104 specifies the preference corresponding to the attribute of the SNS user specified as the estimation partner from the storage content of the storage unit 103 (Step S16).

  The presentation unit 105 presents an item corresponding to the preference specified for the SNS user as a recommendation partner to the user (step S17).

  According to the above embodiment, the relationship between the attribute and the preference of the SNS user is estimated from a plurality of posting histories, and an item corresponding to the preference of a specific customer user is presented using the estimation result.

<Modification>
The following modifications may be implemented in combination with each other.
<Modification 1>
When estimating the preference of the SNS user from the posting history that matches the predetermined condition, the preference estimation unit 102 performs weighting on the relationship between the posting history and the preference to estimate the preference of the SNS user. Is also good. For example, the following weighting example can be considered.

  For example, the preference estimating unit 102 may determine whether the posted poster is not the original creator of the post when the posted poster satisfies the condition that the posted poster is not the original creator of the post. The weight for the relationship between the user and the preference may be reduced. This assumes that, for example, a post called a tweet of a first SNS user who is a recommendation partner is posted and posted by a second SNS user in a format called a retweet. In this case, since the content posted by the second SNS user may not express the preference of the second SNS user, the posting history of the second SNS user and the second SNS user The weight on the relationship with the preference is reduced. Methods for reducing the weight include multiplying the number of posts at the time of aggregation in statistical processing by a value less than 1, multiplying a value less than 1 when converting a positive or negative score in sentiment analysis, and learning data. For example, a method of downsampling from is possible. Accordingly, when the storage unit 103 stores the attribute estimated by the attribute estimation unit 101 and the preference estimated by the preference estimation unit 102 in association with each other, the relationship between the attribute and the preference is weakened.

  Further, for example, when the condition that the posting is for advertisement or promotion is satisfied, the preference estimation unit 102 may assign a weight to the relationship between the posting history and the preference compared to the posting for purposes other than advertisement or promotion. May be reduced. Also in this case, it is considered that the contents posted for the purpose of advertisement or promotion may not express the taste of the SNS user who is the poster. The method for reducing the weight is the same as in the above example.

<Modification 2>
The relationship between the attribute of the SNS user and the preference of the SNS user may be different depending on the attribute of the SNS user. For example, the degree of the attribute of the occupation type may be smaller than the degree of the attribute of the attribute gender. Therefore, assuming that the weight of the preference corresponding to a certain type of attribute is 1, the preference estimation unit 102 sets the weight of the preference corresponding to another type of attribute to less than 1. That is, the preference estimation unit 102 changes the weight of the relationship between the posting history of the SNS user and the preference according to the attribute of the SNS user estimated by the attribute estimation unit 101. More specifically, the preference estimating unit 102 changes the weight of the relationship between the posting history and the preference of the SNS user according to the type of the SNS user attribute estimated by the attribute estimating unit 101.

  In addition, the estimation accuracy when the attribute estimating unit 101 estimates the attribute of the SNS user may differ depending on the attribute. For example, the accuracy when the attribute estimating unit 101 estimates the attribute of age (or age) may be higher than the accuracy when the attribute estimating unit 101 estimates the attribute of occupation. Therefore, when the age and the occupation are estimated as the attributes of the customer user as the recommendation partner, the preference estimating unit 102 assumes that the weight of the preference corresponding to the attribute of the age is 1, and the weight of the preference corresponding to the attribute of the occupation. To less than 1. That is, the preference estimation unit 102 changes the weight of the relationship between the posting history of the SNS user and the preference according to the attribute of the SNS user estimated by the attribute estimation unit 101. More specifically, the preference estimation unit 102 changes the weight for the relationship between the posting history and the preference of the SNS user according to the estimation accuracy of the attribute of the SNS user estimated by the attribute estimation unit 101.

  Further, as described above, there is a possibility that the attribute estimation by the attribute estimating unit 101 may be erroneously determined. Therefore, the preference estimating unit 102 considers attributes other than the estimated attributes, and The weight for the relationship with the preference may be changed. For example, if the estimation accuracy of a student as an attribute is 70%, the possibility of 100-70 = 30% for other attributes (eg, office worker, housewife, self-employed) is evenly determined. Allocate 10% for office workers, 10% for housewives, and 10% for self-employment. That is, the preference estimating unit 102 determines that the recommendation partner is a student with a 70% possibility, a company employee with a 10% possibility, a housewife with a 10% possibility, and a self-wife with a 10% possibility. Assuming that it is a business, its taste is estimated.

  If the probability of the erroneous determination is known, the preference estimating unit 102 changes the weight of the relationship between the posting history of the SNS user and the preference in consideration of the probability of being an attribute other than the estimated attribute. It may be. For example, if the estimation accuracy of a student is 70% as an attribute, the estimation accuracy of a company employee is 15%, the estimation accuracy of a housewife is 10%, and the estimation accuracy of a self-employed person is 50%, The identification unit 104 is that the recommendation partner is a student with a possibility of 70%, a company employee with a possibility of 15%, a housewife with a possibility of 10%, and a self-employment with a possibility of 5%. , The preference is estimated.

<Modification 3>
Even if the attribute exists on the customer list listing the customer users who can be recommended partners, there may be an attribute that has not been estimated from the posting site. For example, there is a case where the customer list has an attribute of 40s, but the attributes of the user estimated from the posting history of a certain item are only teens, 20s, and 30s. As described above, when the storage unit 103 does not store the same attribute as the attribute of the customer user of the recommendation partner, the specifying unit 104 uses the preference of the user of another attribute to set the preference of the customer user to be the recommendation partner. May be specified. For example, the specifying unit 104 specifies the preference of the recommendation partner according to the preference of the customer user whose certain first attribute does not match but the other second attribute of the attribute group of the customer user of the recommendation partner matches. I do. For example, if the attribute of the recommended partner is not a housewife in the forties, but the attribute of the housewives in the twenties and the housewives in the thirties is present in the storage contents of the storage unit 103, the preference corresponding to the attribute of the housewives in the twenties is present. And the average preference of the preference corresponding to the attribute of the housewife in the thirties is specified as an item recommended to the user of the attribute of the housewife in the forties who is a recommendation partner. In this case, the first attribute is “forties” and the second attribute is “housewife”.

  In addition, the specifying unit 104 specifies the preference of the first user to be recommended according to the preference of the customer user of the second attribute having a preference similar to the preference of the first customer user to be recommended. You may do so. For example, on the premise that the preference of the user in the forties is close to the preference of the user in the thirties, the specifying unit 104 determines the recommendation partner based on the preference corresponding to the attribute of the housewife in the thirties, and determines the preference of the customer user having the attribute of the housewife in the forties. Identify preferences.

  Here, for example, a method of determining what attribute a customer user having a preference similar to the preference of a user as a recommendation partner is as follows. First, there is a case where it is determined in advance based on prior knowledge. For example, a preference for a certain fashion is that a housewife and a self-employed person are closer than a housewife and a company employee. That is, the attribute of the customer user having a preference similar to the preference of the customer user as the recommendation partner is determined in advance.

  In addition, there is a method of determining the numerical value of the attribute in the coordinate space based on the proximity of the position. For example, the attribute of residence is such that Kanto and Tokai are closer to Kanto and Kansai, and the attribute of age is that those in their 40s and 30s are closer to their 40s and 20s. That is, the attribute of the customer user having a preference similar to the preference of the customer user as the recommendation partner is determined according to the numerical value of the attribute or the proximity of the position.

  There is also a method of making a determination based on the preference for an item. There is no posting of the SNS user in his forties with respect to the item A, but the SNS user in his forties and the SNS user in his thirties are close to the item B and the item C, so the SNS user in his forties has no post. For example, the preference of the SNS user is specified based on the preference of the SNS user in their thirties. That is, the preference of the SNS user as the recommendation partner with respect to the first item is determined according to the preference of another user with respect to the second item.

The present invention can also be provided in the form of a program for causing a computer to function as an information processing device, or a recording medium on which such a program is recorded. Further, the program according to the present invention may be downloaded to a computer via a network such as the Internet.
Further, the image processing apparatus according to the present invention may be built in the image forming apparatus.

DESCRIPTION OF SYMBOLS 1 ... Information processing apparatus, 2a, 2b ... Contribution site, 3 ... Network, 10 ... Information processing system, 11 ... Control part, 12 ... Storage part, 13 ... Communication Unit 100 acquisition unit 101 attribute estimation unit 102 preference estimation unit 103 storage unit 104 identification unit 105 presentation unit.

Claims (15)

Attribute estimating means for estimating the attribute of each user from post histories of a plurality of users;
Preference estimation means for estimating the preference of each user from the posting histories of the plurality of users;
Storage means for storing the estimated attribute and the preference in association with each other,
Specifying means for specifying the preference corresponding to the attribute of the user to be recommended from the storage content of the storage means;
An information processing apparatus, comprising: a presentation unit configured to present an item corresponding to a preference specified for the user as the recommendation partner to the user as the recommendation partner. The preference estimating means estimates the preference of the user by weighting the relationship between the posting history and the preference when estimating the preference from the posting history that matches a predetermined condition. The information processing device according to claim 1. The preference estimating means, when the posting user is not the original creator of the posting history, compared with the case where the posting user is the original creator of the posting history, the posting history and the preference The information processing apparatus according to claim 2, wherein a weight for the relationship is reduced. The preference estimating means reduces the weight of the relationship between the posting history and the preference with respect to the posting history for the purpose of advertisement or promotion, as compared to the posting for the purpose other than advertising or promotion. The information processing device according to claim 2. The information processing apparatus according to claim 1, wherein the preference estimating unit changes a weight of a relationship between the posting history of the user and the preference according to an attribute of the user estimated by the attribute estimating unit. The information processing apparatus according to claim 5, wherein the preference estimating means changes a weight for a relationship between the posting history of the user and the preference according to a type of an attribute of the user estimated by the attribute estimating means. apparatus. 6. The information according to claim 5, wherein the preference estimating unit changes a weight for a relationship between the posting history of the user and the preference according to estimation accuracy of a user attribute estimated by the attribute estimating unit. 7. Processing equipment. 8. The information processing apparatus according to claim 7, wherein the preference estimating unit changes a weight for a relationship between the posting history of the user and the preference in consideration of an attribute other than an attribute of the user as the recommendation partner. 9. . The said preference estimation means changes the weight with respect to the relationship between the posting log | history of the said user and the said preference in consideration of the probability which is an attribute other than the attribute estimated by the said attribute estimation means. Information processing device. When the same attribute as the attribute of the user to be recommended is not stored in the storage unit, the specifying unit specifies the preference of the user to be the recommended partner using the preference of the user of another attribute. The information processing apparatus according to claim 1, wherein: When the storage unit does not store the same first attribute as the attribute of the user to be recommended, the specifying unit may include a preference corresponding to a second attribute of the attribute of the user to be recommended. The information processing apparatus according to claim 10, wherein If the same attribute as the first attribute of the user to be recommended is not stored in the storage unit, a second attribute having a preference similar to the preference of the user to be recommended is provided. The information processing apparatus according to claim 10, wherein a preference corresponding to the same second attribute as the attribute of the user to be recommended is specified in accordance with the preference of the user. 13. The information processing apparatus according to claim 12, wherein attributes of a user having a preference similar to the preference of the user as the recommendation partner are determined in advance. The information processing apparatus according to claim 12, wherein an attribute of a user having a preference similar to the preference of the user as the recommendation partner is determined according to a numerical value of the attribute or proximity of a position. The information processing apparatus according to claim 12, wherein the preference of the user as the recommendation partner with respect to the first item is determined according to the preference of another user with respect to the second item.

Images